Libs-API Meeting 2023-04-04

tags: Libs Meetings Minutes

Meeting Link: https://meet.jit.si/rust-libs-meeting-crxoz2at8hiccp7b3ixf89qgxfymlbwr
Attendees: Amanieu, David, Mara, The 8472, Urgau

Agenda

• Triage
• Library team meetup
• Triaging ACPs in weekly meetings?
• Anything else?

Triage

(nominated) rust.tf/54140 Tracking Issue: Procedural Macro Diagnostics (RFC 1566)

New design idea (from Esteban and Mara) here: https://github.com/rust-lang/rust/issues/54140#issuecomment-1487529341

Open question: silencing warnings/notes? Some thoughts here: https://github.com/rust-lang/rust/issues/54140#issuecomment-1488547390

David: clippy already has many warnings that aren't of the highest quality, what should we expect from proc macros from the ecosystem?

Do we need a top level "note"? Maybe that should be "trace" or "debug" if that's how it's intended to be used?

Main reason people want this today is because of warnings/notes, as errors are already possible with a hack. So only stabilizing error() doesn't help that much.

'handle' style API seems good. Only open question is the lint/suppression question.

David: the lint!(just_the_name) idea could work. Documentation could just be a doc comment on the static item.

Open question on how to namespace lints from proc macros. E.g. #[allow(clippy::blah)] doesn't refer to the create "clippy".

(nominated) rust.tf/109660 Document that SystemTime does not count leap seconds

This is true on most (all?) operating systems, but do we want to guarantee that?

We do define it as a unix timestamp, which doesn't include leap seconds..

Converting a SystemTime to an actual date+time means subtracting UNIX_EPOCH and interpreting the Duration, assuming no leap seconds. So crates already assume this.

SystemTimes can be created for the future, and we don't know the future leap seconds yet.

no leap seconds.

(nominated) rust.tf/rfc3240 Edition Based Method Disambiguation: Preventing inference ambiguity breakages with extension trait methods

Postponed until the in-person meetup.

(nominated) rust.tf/rfc3362 Command improvements for ergonomics and error handling

No updates. To be discussed in the in-person meetup.

(nominated) rust.tf/rfc3365 RFC: re-export stdlib macros from submodules

Discussed last time. Posted the conclusion an hour ago.

(waiting-on-team) rust.tf/99262 Tracking Issue for `io_error_downcast`

Just needs a @bors rfc merge?

David to check the thread before starting FCP.

Triaging ACPs in weekly meetings?

We don't have a good process for closing an ACP. And we don't define a time for the 'initial commenting period'.

Let's update docs and processes during the in-person meetup.

Maybe just re-use the MCP tooling that the lang+compiler team already use.

FCPs

1 rust-lang/rfcs T-libs-api FCP

• merge rust.tf/rfc3394 format-dynamic-fill - (4 checkboxes left)

11 rust-lang/rust T-libs-api FCPs

• merge rust.tf/107624 Stabilize `const_cstr_methods` - (3 checkboxes left)
• merge rust.tf/88581 Tracking Issue for `int_roundings` - (1 checkboxes left)
• merge rust.tf/80437 Tracking Issue for `box_into_inner` - (1 checkboxes left)
• merge rust.tf/91946 Tracking Issue for `io::Error::other` - (3 checkboxes left)
• merge rust.tf/98461 Document lack of panic safety guarantees of `Clone::clone_from` - (0 checkboxes left)
• merge rust.tf/52331 Correcting Path::components on Redox - (5 checkboxes left)
• merge rust.tf/82901 Tracking Issue for `Option::get_or_insert_default` - (2 checkboxes left)
• merge rust.tf/83871 Tracking Issue for CharIndices::offset function - (4 checkboxes left)
• close rust.tf/51540 Tracking issue for the #[alloc_error_handler] attribute (for no_std + liballoc) - (3 checkboxes left)
• merge rust.tf/88989 Tracking Issue for #![feature(unix_chown)] - (4 checkboxes left)
• merge rust.tf/102341 Implement `Neg` for signed non-zero integers. - (3 checkboxes left)

1 rust-lang/stdarch T-libs-api FCP

• merge github.com/rust-lang/stdarch/issues/1399 Stabilize AArch64 AES/SHA1/SHA2 intrinsics - (3 checkboxes left)